Algorithmic Simplicity and Relevance

TL;DR

This paper proposes a model where human relevance judgments are based on the simplicity of describing situations, linking cognitive processes to complexity measures to predict interestingness and conversational moves.

Contribution

It introduces a novel complexity-based framework for understanding relevance and decision-making in human cognition and communication.

Findings

Situations are deemed relevant when simpler to describe than to generate.

The model predicts interestingness based on descriptive simplicity.

It offers a falsifiable approach to studying relevance in cognition.

Abstract

The human mind is known to be sensitive to complexity. For instance, the visual system reconstructs hidden parts of objects following a principle of maximum simplicity. We suggest here that higher cognitive processes, such as the selection of relevant situations, are sensitive to variations of complexity. Situations are relevant to human beings when they appear simpler to describe than to generate. This definition offers a predictive (i.e. falsifiable) model for the selection of situations worth reporting (interestingness) and for what individuals consider an appropriate move in conversation.